best thermal conductivity in soldering/brazing

[Robert1998]

Hi everyone.

I want to make a heat sink attaching a copper tube to a copper plate (100x100mm, 1mm thick). I´ll try to flat the tube to increase the surface contact but I´ll see.

The question is, to fix the tube to the plate I´m thinking in soldering with Sn92 ag8 or brazing it with copper rods or copper/silver(40%) rods.

I think is better brazing because the rods are made of metals with hight thermal conductivity but, because the hight temperature, Will the plate bend? It is important to remain flat.

Any advice about the metod or the materials used?

Thank you

best regards

 

[Wildefalcon]

Could you cut the plate 120x100 and cut the excess to make a tag/tab at one side? Less risk of distortion.

 

[Ross365]

What's the diameter of the tube? What is the best thermal conductivity value you have that you'd like to beat? I've got on to this previously making heat-exchange plates for cells used in chemical spectrometers. Our requirement was easy to meet and we used lead-tin solder on the 6mm o.d. tube, but I have a possible thought I need to check out.

 

[Kram]

Whats the application, how much power will be applied and what size is the heat source and max temp? Are you relying on air convection or water cooling with the pipe?

The braze joint is an additional thermal interface but if its thickness can be minimised, it will have less effect. If both parts are flat and in good contact then Id use a high flow rod like silverflo 55. If its not flat or radiused pipe joined to a flat, heat transfer will be much lower.

For a 100mm² plate, 1mm thickness will not provide great heat spreading.
Ideally the surface would be lapped or machined flat after manufacture. Brazing temperatures will also anneal copper and at 1mm thick it will be very easily bent when fitting.


Chances are, theres an off the shelf part that will do a much better job and much cheaper than you can make it.

 

[Robert1998]

Thanks for your answers. I know I can buy it, but I enjoy making this things. I have no ideal valure, just get the best with the materials I have. I made a msitake, the plate is not 1mm but 1,5 mm. But it has to be 100x100.

The pipe is round but I´m going to try to falt it a little.

The aplications is as a heatsink of 4 peltier cells.

And the copper brazing rods aren´t 100% cooper, it contaisn a bit of phosporus and silver so the melting point is 645 ºC

 

[frank horton]

bearing mind the melting point of solder, I'd be going down that route....
not sure which has the highest melt point, electrical or plumbers solder.....
I have several large soldering irons....actually beasts.....one u heat up on the cooker and some electricaly.....
the contain enugh heat to resolder car radiator hose connections....
plus theres an epoxy that will do the same job, I have not used it as it's very very expensive...

 

[Ross365]

The thread starts by saying it would be best to using joining materials with high thermal conductivity but where are the numbers on the alternative materials??? Without that data, this is a pointless discussion.

 

[DAPPH]

Just googled the name of the cells

Lots of heat sink s tyles come to mind including using thermally conductive paste / glue to stick them to bespoke hand made copper heatsinks .
What sort of mounting arrangement will the cells be made to fit at the circuit board dimensions & spacings etc ? How much space is available for each one ?

 

[Kram]

The thread starts by saying it would be best to using joining materials with high thermal conductivity but where are the numbers on the alternative materials??? Without that data, this is a pointless discussion.

Its only a small part of the design that if sdone well shouldnt have any significant impact on performance, but there is a lot to be considered, like how will it attach? Mechanical clamping will likely bendtthe copper.

Im going to assume 4x 40mm², 40W TEC's, and if I remember correctely that 160w on cold side will double to 320w on the hot side. You get about 60c max difference between the hot and cold. If you cant keep the hot side temperatures down the cold side will cook.

And the copper brazing rods aren´t 100% cooper, it contaisn a bit of phosporus and silver so the melting point is 645 ºC

Sounds like JM Silbralloy, good stuff, I have used it often.

 

[Robert1998]

Thank you for your answers. Sorry for give so less information. And my english is not the best to explain myself.

I´ll going to attach the 4 peltiers, hot side to a 100x100mm copper plate. I have changed from 1.5 to 4mm thickness folowing the advice in this forum. In the other side I´m going to braze with a silver alloy a spiral made of copper tube. I´ll try to flat the tube a bit to increase the surface of contact. The tube will be 5mm OD. If I run cold water (by this tube, I think I culd get a good temperature in the cold side.